cor_make: Make mircoRNA feature correlations

In MahShaaban/sqlome: Build SQLite tables of microRNAs and Transcription Factors-gene Correlations

Description

Calculate the pearson's correlation between microRNA and feature (gene/protein) matrices. The function calls the R base cor function. More importantly the function prepare the input matrices before the call. First, if matches the col.names/bcr. Second, remove duplicated entries. Finally, it removes the microRNAs and features with expression values less than one. In addation, The function returns a tidy data.frame rather than the raw matrix output of cor.

Usage

cor_make(mi, m, cohort, tidy = TRUE)

Arguments

mi	A tidy matrix of microRNA expression data.
m	A tidy matrix of gene or protein expression data.
cohort	A charachter string of the TCGA ID.
tidy	A logical

Value

A tidy data.frame of four columns; miRBase ID, feature id, correlation and TCGA study name.

Examples

# load required libraries
library(RTCGA.rnaseq)
library(RTCGA.miRNASeq)
library(RTCGA.RPPA)

# calculate correlation for genes
ACC.mi <- mirna_tidy(ACC.miRNASeq)
ACC.m <- mrna_tidy(ACC.rnaseq)
corr <- cor_make(ACC.mi, ACC.m, 'ACC')

# calculate correlation for proteins
ACC.mi <- mirna_tidy(ACC.miRNASeq)
ACC.r <- rppa_tidy(ACC.RPPA)
corr <- cor_make(ACC.mi, ACC.r, 'ACC')

MahShaaban/sqlome documentation built on May 28, 2019, 1:46 p.m.